The open-source voice AI studio.
Your voice data shouldn't leave your machine. Vocally AI is home to voice products that run inside your boundary — on your Mac, on your server, in your browser. Bring your own model. No accounts. No credits. No tracking. No cloud dependency.
- 🔒 Privacy is the product — recordings and transcripts stay on your machine; nothing leaves unless you send it
- 🧠 Bring your own model — Whisper tiny to large-v3, any GGUF, any OpenAI-compatible endpoint; the bundled default is a floor, not a ceiling
- 📦 Open source, shipped — MIT-licensed and usable by strangers, not just readable
This isn't one person's shelf. If you build open-source voice software that respects the boundary — permissive license, local-first or self-hostable, bring-your-own-model, nothing phones home, and shipped — there's a slot here for it. Submissions open once the first studio products are public.